Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The culture of rejection in computer science publications

sigbed.org

Post reply on HN
Print view View original

91–100 of 172 posts

Re: The culture of rejection in computer science publications

#91

Post by g9yuayon »

krinchan wrote: post #39

The academic system is completely broken for Computer Science, and I don't really see a way to fix it. The economic realities of the field just make it too risky to allow an exceptionally gifted individual to remain out in the open publishing research that could potentially destroy your business model. Justifying these low acceptance rates as somehow prestigious is really just creating even more perverse incentives t…

I saw a list of acceptance rate here: https://github.com/lixin4ever/Conference-Acceptance-Rate.

Is 18% or so acceptance rate really low, though? Almost 2 in 10 submissions are accepted, and I thought "the top" meant something like 2% or less.

BTW, is there any resources that catalogs which ideas in papers may work well in industry? As someone outside of academia, I find there are simply too many papers, even from top conferences, for me to consume. It's hard for me to know which paper's ideas can help me or not, and this 18% acceptance rate is not a good enough filter any more.

Re: The culture of rejection in computer science publications

#92

Post by ghaff »

chasil wrote: post #82
krinchan wrote: post #39

The academic system is completely broken for Computer Science, and I don't really see a way to fix it. The economic realities of the field just make it too risky to allow an exceptionally gifted individual to remain out in the open publishing research that could potentially destroy your business model. Justifying these low acceptance rates as somehow prestigious is really just creating even more perverse incentives t…

I have never pursued this type of publication, but why on earth does an "acceptance rate" even exist for these (journal or conference) publications? Why not publish them all? Endorse those that are selected, add commentary to those with which there is disagreement, but is a batch inclusion of them all so technically difficult?

Because there is an implied stamp of approval for publications.

These days nothing is keeping people from publishing elsewhere if they want to.

Re: The culture of rejection in computer science publications

#93

Post by BeetleB »

qzx_pierri wrote: post #19

It doesn't matter what color you are. It doesn't matter where you're from. If you create something like what Dijkstra created back in 56' (undeniable innovation), then no door will remain shut. That's the beauty of Math, CompSci, or STEM as a whole. Not to be too blunt, but I'd rather we keep this soft, "everyone should be accepted" nonsense out of CompSci. The cream will rise to the top. Anyone complaining just need…

> The cream will rise to the top.

What makes you think so?

Having been in academia, my experience doesn't match this sentiment.

Re: The culture of rejection in computer science publications

#95

Post by mikkergp »

qzx_pierri wrote: post #56
mikkergp wrote: post #23

Earlier quoted context omitted.

> If you create something like what Dijkstra created back in 56' (undeniable innovation), then no door will remain shut. That's the beauty of Math, CompSci, or STEM as a whole. Don’t get me wrong, this seems like a nice idea, I’d love to live in this world, but hasn’t it been disproven time and again throughout history? This board itself is flooded with engineers with great ideas being dismissed because of “realities…

> This board itself is flooded with engineers with great ideas being dismissed because of “realities of the business” or “changing priorities”, much less “political realities” etc Your argument would be rock solid if those great ideas were only applicable/useful in huge bureaucratic organizations (which is the type of environment where bullshit ideas are given more attention than truly innovative and impactful ideas)…

> However, the next step would be to go where your ideas can be heard. And if the idea is good enough, people will listen

I don't see how your point is in conflict with my point. You originally said that "no door will remain shut". Your response seems to be similar to my point is that this is not in fact true, ideas have to be marketed to the people who need them or "the right doors". It is not in fact "the best ideas that will float to the top" by some objective measure but the most relevant/politically applicable/subjectively best to the audience" ideas that will float to the top.

Re: The culture of rejection in computer science publications

#96

Post by nautilius »

MontyCarloHall wrote: post #34
nautilius wrote: post #28

Earlier quoted context omitted.

Haha, yeah no way anyone’s going to game that system! No way some clique recruits all of the voting power (collaborators to undergrads) to push their work through, all it takes is LeCun to simply read all papers and his vote will balance everything out!

Easy solution: make it so that voting costs karma, so individuals only have limited voting power.

Easy fix to your solution: get more people to vote. How about all 10000 undergrads in your institution?

In your model, a modern day Einstein would have never had a chance; industrialized research, with publication mills and systematic co-authorship permutation is exactly what would flourish under your model.

Re: The culture of rejection in computer science publications

#97

Post by MontyCarloHall »

nautilius wrote: post #96
MontyCarloHall wrote: post #34

Earlier quoted context omitted.

Easy solution: make it so that voting costs karma, so individuals only have limited voting power.

Easy fix to your solution: get more people to vote. How about all 10000 undergrads in your institution? In your model, a modern day Einstein would have never had a chance; industrialized research, with publication mills and systematic co-authorship permutation is exactly what would flourish under your model.

10k undergrads would have few to no highly upvoted publications, and thus little to no voting power, even in aggregate. An even smaller fraction of them would have highly upvoted papers in the specific subfield of the paper they’re voting on, making their aggregated voting power even more worthless.

The only way to game the system would be to somehow convince a ton of highly upvoted people in a given subfield to upvote a paper in that particular subfield (and give up some of their voting power accordingly, since voting would cost you karma). Heck, make it so that you can only vote on a certain number of papers each month (5-10?), to make votes even more valuable.

Re: The culture of rejection in computer science publications

#98

Post by chasil »

ghaff wrote: post #92
chasil wrote: post #82

Earlier quoted context omitted.

I have never pursued this type of publication, but why on earth does an "acceptance rate" even exist for these (journal or conference) publications? Why not publish them all? Endorse those that are selected, add commentary to those with which there is disagreement, but is a batch inclusion of them all so technically difficult?

Because there is an implied stamp of approval for publications. These days nothing is keeping people from publishing elsewhere if they want to.

If the organization is receiving public funds of any kind, then they should be required to (electronically) publish all submissions unless the authors withdraw.

There has been quite enough censorship and paywalling of research that my taxes fund.

Re: The culture of rejection in computer science publications

#99

Post by 22SAS »

faangiq wrote: post #94

CS has a ton of people in it and they produce a lot of garbage papers. Oh you retrained your ML model? Boom, new paper.

I completely agree with this. I have seen papers produced by a bunch of newly minted PhD's and the first thought that came in my mind was "WTF is this? There's nothing in this that's new or makes sense".

A lot of publishing, from outside the top research universities, seems to be mainly at showing a higher publication count to get more funding and impress tenure committees.

Some of the professors, who do this, make you wonder how they get their PhD in the first place. Back in my masters, I was giving a presentation on a paper from FB, where they were going over scheduling policies they had been using in their hadoop clusters. The professor reviewing asked me a question "You're taking about scheduling tasks in distributed environments, you're talking about Hadoop, Hadoop is for files what scheduling does it do?". Some of us there had a big "WTF" plastered on our face. Couldn't even argue cause this prof. thought he was a hotshot and would make life hell for students who did. In his own words to a student "I can write a compiler, I can write an operating system, but I can't make head or tail of what you're doing".

Re: The culture of rejection in computer science publications

#100

Post by walleeee »

lioeters wrote: post #43
LunaSea wrote: post #16

Earlier quoted context omitted.

Could you give a bit more context about this? I'm not familiar with the subject.

I think they're just joking that some of the material presented at Chinese academic conferences are close to non-sense, might as well be produced by DALL-E.

Can't speak to the state of Chinese academia but have had a few similar experiences at American conferences

When careers are on the line and one is expected to mischaracterize one's work or its impact, people will tend to do so

If this problem exists independently of cultural or institutional factors maybe there are potential solutions with an equally broad range of application

Post reply on HN